On Tuesday 17 December the UK government launched an industry consultation on reform to copyright law in relation to artificial intelligence (AI). 

The launch of the consultation is a notable development after plans for a new AI copyright code were shelved by the previous government in February 2024. The concept of an AI copyright code had itself been put forward after plans for AI copyright deregulation proposed in March 2023 were met with widespread criticism from stakeholders in the creative industries.  

The new consultation seeks views on how the government can deliver reform that achieves 3 key objectives for the AI sector and creative industries:

  1. Supporting control of rights holders’ content and remuneration;
  2. Continuing to develop world-class AI models in the UK through legally-accessed data; and 
  3. Promoting trust and transparency between sectors.

With these objectives in mind, the consultation seeks input on a range of issues affecting the intersection of generative AI on the one hand and IP rights on the other, particularly copyright. It covers issues such as labeling of gen AI content, collective licensing, copyright protection for computer generated works, and digital replicas (aka deepfakes). See our previous analysis of many of these issues under the current legal framework here)

The issue at the heart of the consultation is the regulation of text and data mining of copyright works. Under current copyright law TDM is permitted only for non-commercial research purposes – not for training an AI model intended for commercial exploitation, for example. 

On TDM, the consultation proposes a mechanism for rights-holders to reserve their rights, "enabling them to license and be paid for the use of their work in AI training". This approach is described as "balancing rights holders' ability to seek remuneration while providing a clear legal basis for AI training with copyright material".

Notably, the proposed mechanism appears to depart from the approach proposed by the previous government’s AI Regulation White Paper, presented by Sir Patrick Vallance in March 2023. The White Paper had proposed a blanket exemption for TDM for the training of AI models. The consultation’s proposal instead looks to mirror the EU Digital Single Market Copyright Directive (2019/790), which allows rights-holders to ‘opt out’ from the use of their works for TDM purposes.  

The intention is that this will allow the UK to maintain its position as a market leader in both AI and the creative industries, by giving AI developers access to high quality training data while at the same time adequately compensating rights-holders for the use of their copyright works.  

The consultation’s proposals also call for greater transparency from AI developers as a "pre-requisite", covering the material they use to train models, how they obtain such materials, and the content they generate. The consultation seeks views on how best to increase trust in and around developers in these regards and acknowledges that legislation is likely to be necessary, given the speed of AI development. 

The consultation will be open to all industry stakeholders until 25 February 2025. 

This post was written by Harry Jewson and Abigail Cropper. If you would like to discuss any of the issues raised in this article please contact Harry Jewson or Emily Roberts. 

For the latest updates on AI law, regulation, and governance, see our AI blog at: AI: Burges Salmon blog (burges-salmon.com). 

Subscribe to our Concept newsletter and receive the latest intellectual property legal updates, news and event invitations direct to your inbox.